IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 47, NO.
1, FEBRUARY 2005
Uncertainties in 3-m Radiated Emission
Measurements Due to the Use of Different
Types of Receive Antennas
Nico van Dijk
Abstract—This paper describes the results of an investigation
into uncertainties in radiated emission measurement results due
to the use of different receive antennas. These antennas are used
during electromagnetic-compatibility tests of products which
must comply with regulatory radiated emission requirements.
In the past few years, the standardization International Special
Committee for Radio Protection (CISPR) Subcommittee A (SCA)
has tried to improve the radiated emission measurements by
quantifying most of the uncertainty sources. In this paper, we
deal with the uncertainty due to the free space antenna factor’s
inability to fully specify the antenna properties at frequencies
up to 1000 MHz. We have simulated and measured the radiated
emission of an equipment-under-test (EUT) using various types of
receive antennas. The simulated radiated emission measurement
results using a biconical antenna give a maximum difference of
1.2 dB in relation to the use of a tuned dipole. The use of bow-tie
antennas leads to a radiated emission difference of maximal 2.5 dB
in relation to the tuned dipole. Finally, the use of a log-periodical
antenna leads to a maximum deviation of 2.8 dB in relation to
the tuned dipole. This means that the uncertainty caused by
using different types of receive antennas in 3-m radiated emission
measurements may be significant, so the presented uncertainty
budgets in CISPR 16-4-2 should therefore be reconsidered and
expanded.
Index Terms—Antenna factor (AF), electromagnetic compat-
ibility (EMC), radiated emission, receive antenna, simulation,
uncertainties, uncertainty budget.
I. FORMULATION OF THE PROBLEM
T HESE DAYS accredited test houses are obliged to derive
the uncertainty budgets of their radiated emission compli-
ance measurements (ISO/IEC 17 025 [14]). The performance
of the receive antenna is an important uncertainty source in
a radiated emission measurement. The uncertainty associated
with the determination and application of the antenna factor
(AF) of the receive antenna is considered to be an important
influencing quantity which can be used to describe the overall
uncertainty contribution of the receive antenna. The CISPR Sub-
committee A International Special Committee for Radio Protec-
tion (CISPR SCA) has devoted a great deal of time and effort to
the uncertainties of receive antennas by setting up uncertainty
budgets and developing validation regulations for antenna cali-
bration test sites. CISPR SCA is also developing different stan-
dardised antenna calibration methods; much of this work is still
Manuscript received June 18, 2003; revised May 13, 2004.
The author is with Philips Digital Systems Laboratories, Eindhoven 5656 AA,
The Netherlands (e-mail: Nico.v.Dijk@philips.com).
Digital Object Identifier 10.1109/TEMC.2004.842112
0018-9375/$20.00 © 2005 IEEE
77
in progress [11]. It is, therefore, important that we gain an in-
sight into the actual uncertainty induced by the receive antenna.
Previous research [3], [4], [12] investigated uncertainties caused
by the equipment-under-test (EUT) and the measurement instru-
mentation, including the test sites themselves. The measurement
instrumentation uncertainties of the radiated emission measure-
ment are described in detail in CISPR 16-4-2 [10]. We can dis-
tinguish two uncertainty sources related to the use of receive
antennas. The first source of receive antenna uncertainty is the
method of antenna calibration and the related calibration un-
certainty. We know that a free-space-determined AF is often
used, while the antennas are used within chambers which have
highly conductive ground planes. The ANSI C 63.5 standard
site method is an alternative to the free-space AF method, as it
can be used to determine AFs including a ground-plane effect
[2], [15], [20]. In addition to the ambiguity of the different AF
definitions, there is a second problem related to the use of AFs.
Several types of antennas are usually used for radiated emission
measurements, but a tuned dipole is prescribed in CISPR 16-1-4
[9], [5], [6], as the CISPR reference dipole.1
In the past, the use of different receive antennas was jus-
tified under the following assumption: if the AFs of different
antennas are accurately known, then the different receive an-
tennas should give the same result in a radiated emission mea-
surement. However, recent research results [3], [13] suggest that
the AF is not always an adequate figure-of-merit to describe the
overall antenna performance in the case of a nonuniform inci-
dent plane-wave configuration, which is always the case in test
chambers containing a ground plane. It is this uncertainty source
that we consider here. In [17], the AF is also the underlying
problem, and is identified as a limited figure-of-merit in elec-
tric- and magnetic-field measurements using different types of
probes. In this paper, we look especially at the uncertainty in
radiated emission results caused by the deficiencies of the AF
as a figure-of-merit. In [12], numerical simulation tools were
shown to be very useful in analyzing and quantifying uncer-
tainty sources in radiated emission measurements. We, there-
fore, used numerical simulations to investigate and quantify the
“receive antenna type” uncertainty source, and we supported the
numerical simulations by measurements. In Sections II and III
of this paper, we briefly describe the method and the simulated
antennas under test. We discuss the AF results in Section IV.
1Use of the CISPR reference dipole as the basic reference is presently under
discussion [11].
78 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 47, NO. 1, FEBRUARY 2005
Fig. 1. Simulation configuration of a radiated emission measurement.
Section V gives the simulated radiated emission results, with
the measured radiated emission results given in Section VI. In
Sections VII and VIII, we discuss the results and give conclu-
sions, respectively.
II. DESCRIPTION OF THE METHOD
Radiated emission measurements are carried out at a test site,
such as a semi-anechoic room (SAR) by measuring the max-
imum field emitted by an EUT at a specified horizontal test dis-
tance of 3 or 10 m and for a specified height range of the receive
antenna (1–4 m). To simplify the test setup for simulation pur-
poses, a fixed length dipole can be used to simulate an EUT, as
shown in Fig. 1. In this way, the actual radiated emission mea-
surement is modeled as a coupling test between a transmit and
receive antenna.
In practice, this type of simplified coupling test is also used to
validate the performance of test sites; this is called a site atten-
uation (SA) test [12]. However, the quantity of interest for us in
this paper is not the SA but the maximum radiated electric field.
It is important to emphasize that the maximum radiated electric
field is not a quantity that can be measured or simulated
directly; it is derived from the maximum received voltage
multiplied by the AF of the receive antenna
(1)
In this paper, we apply the most commonly used AF definition
within the CISPR community (i.e., the free space AF). This
AF can be obtained by illuminating the receive antenna using
a copolarized uniform incident plane wave (Fig. 2). The re-
ceived voltage is then measured. This yields the following
definition of the AF:
(2)
The FEKO software suite [7], [16] by EM Software & Systems
was used to perform both the AF and the radiated emission sim-
ulations. FEKO is based on the method of moments (MoM).
Details about the numerical techniques implemented in FEKO
Fig. 2. Configuration for determining the free-space AF.
can be found in [16], [19]. Fig. 1 depicts a schematic overview
of the simulated emission measurement configuration. This con-
figuration is based on the coupling of a system of two antennas
at a horizontal distance of m as the equivalent radiated
emission measurement. In all of the configurations discussed
in this paper, the transmit ( m) and receive antenna
( – m) have 50- termination. We used the following
four-step procedure to obtain our radiated emission results.
1) Determine the free space AF [(2)] of the various types of
receive antennas.
2) Determine the maximum received voltage over the height
(1 to 4 m, 10-cm step size) and use each of the various
types of receive antennas.
3) Convert the maximum voltage into a maximum electric
field using the AF obtained in step 1 [(1)].
4) Compare the radiated emission results obtained using the
various types of receive antennas.
The radiated emission results were compared in step 4 with
the results obtained by using tuned dipoles (i.e., CISPR refer-
ence dipoles)2 as described in CISPR 16-1-4 [9]. This compar-
ison is discussed in Section V. The “receive antenna type” un-
certainty assessment was performed using two types of EUTs
(i.e., a “large” EUT simulated by a 100-MHz dipole, referred to
as “Dip100 EUT,” and a “small” EUT simulated by a 250-MHz
dipole, referred to as “Dip250 EUT”). These two EUTs were
used to investigate the possible influence of EUT size on the
uncertainty.
III. SIMULATED ANTENNAS UNDER TEST
We considered the following types of antennas in our uncer-
tainty assessment: tuned dipoles, biconical antennas, log-peri-
odical antennas, and bow-tie antennas. The simulated log-peri-
odical receive antenna was based on the log-periodical part of
the Chase bilog antenna (CBL6141), shown in Fig. 3. The tip of
the log-periodical antenna is depicted in Fig. 4, where we can
see the MoM elements (wire segments and triangular surfaces).
We simulated two models of biconical antennas (i.e., a closed
surface model and a wire model shown in Figs. 5 and 6, respec-
tively). The simulated wire model was based on the Eaton 2198
biconical antenna. The closed surface model of the biconical an-
tenna was only used in the AF evaluation (Section IV) and not
in the simulations of the radiated emission. We also investigated
two types of bow-tie antennas which are depicted in Fig. 7.
2See footnote 1.
VAN DIJK: UNCERTAINTIES IN 3-m RADIATED EMISSION MEASUREMENTS
Fig. 3. Simulated log-periodical antenna discretized in wire segments and
triangular surfaces.
Fig. 4. Tip of the log-periodical antenna discretized in wire segments and
triangular surfaces.
Fig. 5. Closed surface model of biconical antenna.
The bow-tie antennas are based on the “bicon” parts of
so-called bilog antennas, which consist of a log-periodical part
and a bow-tie part. For reference purposes, several tuned dipole
antennas (CISPR reference dipoles) were also used as receive
antennas. The two EUTs were simulated using “Dip100 EUT”
( m) and “Dip250 EUT” ( m), both with
radius mm.
79
Fig. 6. Simulated biconical antenna discretized in wire segments.
Fig. 7. Simulated green (left) and purple (right) Bow-tie antennas discretized
in wire segments.
IV. AF RESULTS
A. Frequency Range 20–300 MHz
In this frequency range, we looked at the AF simulations
of the biconical and the bow-tie receive antennas. The sim-
ulated AFs of the investigated receive antennas are depicted
in Fig. 8. The bow-tie antennas show peaks between 200 and
300 MHz. Further investigation using the antenna patterns
80 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 47, NO. 1, FEBRUARY 2005
Fig. 8. Simulated antenna factors of various types of receive antennas.
proved that the antenna patterns of the bow-tie antennas de-
teriorate rapidly above 200 MHz, showing lobes and zeros;
while below 200 MHz, the bow-tie antenna patterns follow the
well-known doughnut shape as shown in Fig. 9. We also carried
out some AF investigations using biconical receive antennas,
where we compared two biconical antenna models: a closed
surface model and a wire model (Fig. 10). The simulated wire
model was based on an existing biconical antenna (Eaton 2198)
which does not include crossbars. We also simulated the same
wire model including the crossbars to find out what effect they
would have, but in this frequency range, we could not discover
any differences. However, narrowband cage resonances could
appear at these frequencies [1], [18]. Alexander et al. [1] state
that a dipole antenna will exhibit multilobing if the antenna
length is above ; this multilobing was not apparent using
a biconical antenna of the same length. However, we found
that the biconical antenna pattern did also deteriorate at an
antenna length above (350 MHz) (i.e., where multilobing
occurs). Multilobing of antennas results in large uncertainties
in the radiated emission results. This is discussed in detail in
Section V.
B. Frequency Range 300–1000 MHz
In this frequency range, we simulated the AFs of the log-peri-
odical receive antenna. The measured and simulated AFs of the
log-periodical antenna at higher frequencies (300–1000 MHz)
are depicted in Fig. 11. We conclude that the difference between
the simulated and measured AF for the log-periodical antenna
is less than 2 dB. It must be noted that the measured AF results
of the log-periodical antenna, shown in Fig. 11, are achieved
including the bow-tie antenna, while the simulated AF results
represent the log-periodical antenna solely (without bow-tie an-
tenna). Above 300 MHz, the log-periodical antenna is active,
however, we have no information about the interaction.
V. SIMULATED RADIATED EMISSION RESULTS
In the previous section, we analyzed the behavior of the AF
of various receive antennas. In this section, we look at the con-
sequences of the different behavior of the receive antennas in
simulated radiated emission results. In Section V-A, we look at
the lower frequency range (20–300 MHz), and in Section V-B,
we look at the higher frequency range (300–1000 MHz). CISPR
actually requires a radiated emission start frequency of 30 MHz.
We used the start frequency of 20 MHz to make it easier to see
effects around 30 MHz.
A. Frequency Range 20–300 MHz
The horizontal and vertical simulation results of the radiated
emission test of the Dip100 EUT and the Dip250 EUT are de-
picted in Figs. 13–16. In these figures, we can see large devia-
tions in radiated emission results obtained using the bow-tie an-
tennas compared to the radiated emission result obtained with
the biconical antenna. These deviations correspond to the peaks
in AFs of the bow-tie antennas between 200 and 300 MHz which
were observed in Fig. 8. The observation of the large deviations
in radiated emission results due to the peaks in AFs leads to the
fundamental guideline concerning the working bandwidth of re-
ceive antennas characterized by their free-space AF and used
in chambers containing a ground plane (e.g., a SAR): a receive
antenna must have its highest gain in the direction for which
the AF was determined [(1)–(2)]. Based on the antenna patterns
(Fig. 9), we can conclude that the bow-tie antennas shall not be
used above 200 MHz. The biconical antenna shall not be used
above 300 MHz. The design bandwidths of practical antennas
used for EMC applications generally already comply with this
guideline. The differences in radiated emission of the horizontal
polarization of a Dip100 EUT and Dip250 EUT in relation to the
radiated emission obtained by using a tuned dipole are shown in
Tables I and II, respectively. The differences in vertical polariza-
tion are shown in Tables III and IV. The biconical antenna shows
a maximum deviation of 1.2 dB in its working bandwidth of
20–300 MHz. In addition, if we define the working bandwidth
of the bow-tie antennas as from 20 to 200 MHz, then a max-
imum deviation is observed of 2.5 dB. However, it should be
emphasized that the bow-tie antennas are almost always imple-
mented in bilog antennas in practice (i.e., in combination with
log-periodical antennas, see next subsection).
B. Frequency Range 300–1000 MHz
Figs. 17 and 18 show the horizontal and vertical simulation
results in the 300–1000 MHz range for the radiated emission
measurement of the Dip100 EUT and the Dip250 EUT, respec-
tively. The deviations due to the use of the log-periodical an-
tenna are listed in Tables V–VIII, where the differences are
shown in relation to the CISPR reference dipole (tuned dipole).
In these tables, we can observe that for horizontal polarization
(Tables V–VI), the difference is smaller than 1.6 dB at frequen-
cies lower than 800 MHz and at 1000 MHz, the difference is
2.8 dB. In addition, in the vertical polarized configuration (Ta-
bles VII–VIII), the differences are smaller than 1.7 dB up to
1000 MHz. The differences become larger above 800 MHz be-
cause of the fact that multilobing starts to develop above this
frequency. At a frequency of 1000 MHz, the maximum gain of
the log-periodical antenna is not exclusively found at
VAN DIJK: UNCERTAINTIES IN 3-m RADIATED EMISSION MEASUREMENTS 81

Fig. 9. Antenna patterns (gain) of the green (left) and purple (right) Bow-tie antennas at 150, 200, and 250 MHz (Fig. 7).

TABLE I
DIFFERENCES IN RADIATED EMISSION OF THE Dip100 EUT (HORIZONTAL POLARIZATION) IN RELATION TO A TUNED DIPOLE

TABLE II
DIFFERENCES IN RADIATED EMISSION OF THE Dip250 EUT (HORIZONTAL POLARIZATION) IN RELATION TO A TUNED DIPOLE

Fig. 10. Simulated antenna factor of various types of biconical antennas. Fig. 11. Antenna factors of the log-periodical antenna (simulated and
measured).

the boresight direction of 90 anymore (i.e., the direction for
which the free-space AF was determined) [see Fig. 1], but also
near 60 and 120 (Fig. 12). Accordingly, above 1000 MHz,
the used log-periodical antenna shall not be used anymore in
radiated emission measurements performed in a semi-anechoic
room (SAR). It must also be noted that the 10-cm height step
of the receive antenna becomes more critical in the range
800–1000 MHz ( cm). In practice, the log-periodical
antennas are often combined with bow-tie antennas. In the
previous subsection, the maximum difference in radiated emis-
sion results measured using the bow-tie antennas was 2.5 dB.
This maximum was found at 100 MHz (Table II). The working
bandwidth of the log-periodical antenna is 200–1000 MHz.
Although we have no information about the interaction between
the log-periodical antenna and bow-tie antenna, a bilog antenna
with the combination of the log-periodical and bow-tie structure
82 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 47, NO. 1, FEBRUARY 2005

TABLE III
DIFFERENCES IN RADIATED EMISSION OF THE Dip100 EUT (VERTICAL POLARIZATION) IN RELATION TO A TUNED DIPOLE

TABLE IV
DIFFERENCES IN RADIATED EMISSION OF THE Dip250 EUT (VERTICAL POLARIZATION) IN RELATION TO A TUNED DIPOLE

TABLE V
DIFFERENCES IN RADIATED EMISSION OF THE Dip100 EUT (HORIZONTAL
POLARIZATION) IN RELATION TO A TUNED DIPOLE

TABLE VI
DIFFERENCES IN RADIATED EMISSION OF THE Dip250 EUT (HORIZONTAL
POLARIZATION) IN RELATION TO A TUNED DIPOLE

Fig. 12. Antenna patterns (gain) of the log-periodical antenna at 200, 300, 900,
and 1000 MHz.

can possibly result in similar differences of around 2 dB in the

frequency range 20–200 MHz.
TABLE VII
DIFFERENCES IN RADIATED EMISSION OF THE Dip100 EUT (VERTICAL VI. MEASURED RADIATED EMISSION RESULTS
POLARIZATION) IN RELATION TO A TUNED DIPOLE
The presented “receive antenna type” uncertainty assessment
is based on simulations; it is, therefore, important to validate
the simulations by measurements. The measurement results are
shown in Figs. 19 and 20. If we compare these figures with the
corresponding simulated radiated emission results in Figs. 13,
14, and 17, we can see that the measurement results closely
match the simulation results within 3 dB and, therefore, the
use of simulations for these AF-induced uncertainty investiga-
TABLE VIII
tions is considered a valid approach.
DIFFERENCES IN RADIATED EMISSION OF THE Dip250 EUT (VERTICAL
POLARIZATION) IN RELATION TO A TUNED DIPOLE VII. DISCUSSION AND OUTLOOK
In Section IV, we observed that all types of antennas have
a certain frequency above which the antenna patterns start to
deteriorate (i.e., the antenna patterns exhibit multilobing). This
observation led us (in Section V) to formulate a guideline for
the working bandwidths of antennas (i.e., that a receive an-
tenna must have its maximum gain in the direction for which
VAN DIJK: UNCERTAINTIES IN 3-m RADIATED EMISSION MEASUREMENTS 83

Fig. 13. Simulated radiated emission of Dip100 EUT (horizontal polarization) Fig. 16. Simulated radiated emission of Dip250 EUT (vertical polarization)
with various types of receive antennas. with various types of receive antenna.

Fig. 14. Simulated radiated emission of Dip100 EUT (vertical polarization)

with various types of receive antennas.
+
Fig. 17. Simulated radiated emission of Dip100 EUT (vertical horizontal
polarization) measured using a log-periodical antenna.

Fig. 15. Simulated radiated emission of Dip250 EUT (horizontal polarization) +

Fig. 18. Simulated radiated emission of Dip250 EUT (vertical horizontal
with various types of receive antennas. polarization) measured using a log-periodical antenna.

the AF was determined). In the working bandwidths of the an- types of antennas. In the working bandwidths of the investi-
tennas, the assumption stated in Section I, justifying the use gated antennas, a considerable “receive antenna type” devia-
of different antennas was investigated by looking at the devi- tion was found of around 2 dB. The expanded uncertainty ta-
ations in radiated emission results due to the use of various bles presented in CISPR 16-4-2 [10] list the following uncer-
84 IEEE TRANSACTIONS ON ELECTROMAGNETIC COMPATIBILITY, VOL. 47, NO. 1, FEBRUARY 2005

tainty analyses should also include the 10-m configuration. We

expect a smaller “receive antenna type” uncertainty at 10-m test
distance because the deviations in the radiation patterns of the
receive antennas are smaller due to the smaller elevation angles.

VIII. CONCLUSION
In Section I, we questioned whether the AF adequately de-
fines the antenna behavior since differences (up to 3 dB) had
been observed in radiated emission results [3]. In its working
bandwidth of 20–300 MHz, the biconical antenna shows a radi-
ated emission difference in relation to a tuned dipole of maximal
1.2 dB. We had concluded that the working bandwidth of the
bow-tie antennas is 20–200 MHz. In this bandwidth, the bow-tie
antennas exhibit a radiated emission difference in relation to the
tuned dipole of maximal 2.5 dB. The log-periodical antenna ex-
+
Fig. 19. Measured radiated emission of Dip100 EUT (vertical horizontal
hibits in its working bandwidth (200–1000 MHz), a maximum
polarization) measured using a biconical antenna.
deviation of 2.8 dB in relation to the tuned dipole. These devi-
ations, which were achieved using a 3-m test distance, confirm
the “receive antenna type” uncertainty suggested in [3]. This
means that the uncertainty caused by using different types of
receive antennas in radiated emission measurements may be sig-
nificant, so the presented uncertainty budgets in CISPR 16-4-2
[10] should, therefore, be reconsidered and expanded.

ACKNOWLEDGMENT
The author would like to thank P. A. Beeckman of the PDSL
EMC Group at Philips Electronics and member of CISPR SCA
for supervising and supporting this research. Thanks are due to
S. B. Worm and M. J. C. M. van Doorn of Philips Electronics
(PDSL EMC ) for reviewing this paper.

+
Fig. 20. Measured radiated emission of Dip100 EUT (vertical horizontal
polarization) measured using a log-periodical antenna.
tainty quantities related to the antenna: the AF themselves, AF
frequency interpolation, AF height deviation, directivity differ-
ence, phase center location, cross-polarization, and balance. The
budgets presented in CISPR 16-4-2 [10] should, therefore, be
reconsidered and expanded by including the “receive antenna
type” uncertainty source. The receive antenna-type deviations
found using the large and small EUT (Dip100 EUT and Dip250
EUT) exhibit no significant differences. This means that EUT
size in the class of EUTs with the cable as dominating radi-
ator is not of large influence. Future uncertainty analyses should
include the broadband double-ridged horn antenna; especially
above 1 GHz. Bruns et al. [7] state that the commonly used
1-18 GHz broadband double-ridged horn antenna exhibits mul-
tilobing above 12 GHz, so we can expect it to result in con-
siderable radiated emission uncertainty. In the future, it is also
important to investigate the “receive antenna type” uncertainty
if the AF is determined using the ANSI C 63.5 standard site
method [20]. Finally, in this paper, we have investigated the un-
certainty due to the use of different types of receive antennas
in a 3-m SAR test site. However, radiated emission measure-
ments are often performed in 10-m test sites. So, future uncer-
REFERENCES
[1] M. J. Alexander et al., “Getting the best out of biconical antennas
for emission measurements and test site evaluation,” in Proc. IEEE
Int. Symp. Electromagnetic Compatibility, Austin, TX, Aug. 1997, pp.
84–89.
[2] “Calibration of Antennas Used for Radiated Emission Measurements in
Electromagnetic Interference (EMI) Control,” ANSI C63.5, Jun. 1988.
[3] P. A. Beeckman and J. J. Goedbloed, “Results of the CISPR/A radiated
emission round robin test,” in Proc. IEEE Int. Symp. Electromagnetic
Compatibility, vol. 1, Montreal, QC, Canada, Aug. 2001, pp. 475–480.
[4] P. A. Beeckman, “The influence of positioning tables on the results of
radiated EMC measurements,” in Proc. IEEE Int. Symp. Electromagnetic
Compatibility, vol. 1, Montreal, QC, Canada, Aug. 2001, pp. 280–285.
[5] C. E. Brench, “Antenna factor anomalies and their effects on EMC mea-
surements,” in Proc. IEEE Int. Symp. Electromagnetic Compatibility,
Atlanta, GA, Aug. 1987, pp. 342–346.
[6] , “Antenna differences and their influence on radiated emission
measurements,” in Proc. IEEE Int. Symp. Electromagnetic Compati-
bility, Washington, DC, Aug. 1990, pp. 440–443.
[7] C. Bruns et al., “Analysis and simulation of a 1–18-GHz broadband
double-ridged horn antenna,” IEEE Trans. Electromagn. Compat., vol.
45, no. 1, pp. 55–60, Feb. 2003.
[8] “Information Technology Equipment—Radio Disturbance Characteris-
tics—Limits and Methods of Measurement,” CISPR 22, 1997.
[9] “Specification for Radio Disturbance and Immunity Measuring
Apparatus and Methods—Part 1-4: Radio Disturbance and Immunity
Measuring Apparatus—Ancillary Equipment—Radiated Disturbances,”
CISPR 16-1-4, 2003.
[10] “Specification for Radio Disturbance and Immunity Measuring Appa-
ratus and Methods—Part 4-2: Uncertainties, Statistics and Limit Mod-
eling—Uncertainty in EMC Measurements,” CISPR 16-4-2, 2003.
[11] “Questionnaire Antenna Calibration,” CISPR/A/454/Q, 2003.
VAN DIJK: UNCERTAINTIES IN 3-m RADIATED EMISSION MEASUREMENTS
[12] N. van Dijk, “Numerical tools for simulation of radiated emission testing
and its application in uncertainty studies,” IEEE Trans. Electromagn.
Compat., vol. 44, no. 3, pp. 466–470, Aug. 2002.
[13] T. J. Dvorak, “Are antenna factors valid in EMC measurements?,” in
Proc. Int. Symp. Electromagnetic Compatibility, Nagoya, Japan, Sep.
1989, pp. 587–590.
[14] “General Requirements for the Competence of Testing and Calibration
Laboratories,” ISO/IEC 17 025, 1999.
[15] U. Jakobus and F. M. Landstorfer, “Numerical analysis of errors associ-
ated with antenna calibrations and emission testing,” in Proc. 12th Int.
Zurich Symp. Electromagnetic Compatibility, Feb. 1997, pp. 441–446.
[16] U. Jakobus et al.. (2002) FEKO—Field computation involving objects of
arbitrary shape, Stellenbosch Univ., South Africa. [Online]. Available:
www.feko-usa.com
[17] W. Joseph and L. Martens, “The influence of the measurement probe
on the evaluation of electromagnetic fields,” IEEE Trans. Electromagn.
Compat., vol. 43, no. 2, pp. 339–349, May 2003.
[18] S. M. Mann and A. C. Marvin, “Characteristics of the skeletal bicon-
ical antenna as used for EMC applications,” IEEE Trans. Electromagn.
Compat., vol. 36, no. 4, pp. 322–330, Nov. 1994.
[19] M. N. O. Sadiku, Numerical Techniques in Electromagnetics. Boca
Raton, FL: CRC, 1992.
85
[20] A. A. Smith Jr., “Standard-site method for determining antenna factors,”
IEEE Trans. Electromagn. Compat., vol. EMC-24, no. 3, pp. 316–322,
Aug. 1982.
Nico van Dijk was born in Hengelo, The Nether-
lands, in 1975. He received the polytechnics degree
(Hons.) in electrical engineering from Polytechnical
education in Enschede in 1998 and the M.S. degree
in electrical engineering from the Eindhoven Univer-
sity of Technology, Eindhoven, The Netherlands, in
2001.
Currently, he is with EMC Competence Centre
of Philips Electronics, Eindhoven, The Netherlands.
His main subject was electromagnetic waves and
antennas. His research interests include test methods
(especially above 1 GHz) and future standardization, measurement uncertain-
ties, and reverberation chamber technology.